Sarcocystis spp. are cyst-forming, intracellular protozoan parasites with an obligate two-host life cycle based on a prey-predator relationship. Sexual reproduction occurs in the intestinal epithelial cells of carnivorous definitive hosts, leading to the excretion of oocysts or sporocysts. In herbivorous intermediate hosts, asexual replication results in the formation of intramuscular sarcocysts (Dubey et al. 2016). Sheep (Ovis aries) are intermediate hosts for at least nine Sarcocystis species, including S. tenella (Moulé 1886), S. gigantea (Ashford 1977), S. medusiformis (Collins et al. 1979), S. arieticanis (Heydorn 1985), S. microps (Wang et al. 1988), S. cystiformis (Wang et al. 1989). S. mihoensis (Saito et al. 1997), S. gracilis-like (Giannetto et al. 2005)d mihoensis-like (Gjerde et al. 2020), which are primarily differentiated by their sarcocyst wall ultrastructure. Notably, only S. gigantea and S. medusiformis develop macroscopically visible sarcocysts, informally categorized by their gross morphology as "fat" and "thin" cysts, respectively (Collins et al. 1979). Natural Sarcocystis infections in sheep can induce significant clinical pathology, including weight loss, abortion, myocarditis, encephalitis, and in severe cases, acute mortality (Railliet 1886; Dubey et al. 1989; Scott and Sargison 2001; Yaziroglu and Beyazit 2005). Moreover, the presence of macrocysts formed by S. gigantea and S. medusiformis, along with associated eosinophilic myositis, frequently leads to partial or complete carcass condemnation, incurring substantial economic losses in both lamb and adult sheep production (Collins 1980; Ezzi et al. 1992).

Among the Sarcocystis species infecting sheep, S. tenella, S. arieticanis, and S. gigantea are globally distributed (Dubey et al. 2016; Feng et al. 2023). In contrast, reports of S. medusiformis have been more geographically restricted, with confirmations limited to New Zealand (Collins et al. 1979), Australia (O'Donoghue and Ford 1986; Obendorf and Munday 1987), Iran (Farhang-Pajuh et al. 2014), Iraq (Nawshirwan et al. 2023), Italy (Pipia et al. 2016), Egypt (El-Morsey et al. 2021), and Spain (Gjerde et al. 2020; Peris et al. 2024). This study reports the first detection of S. medusiformis in Chinese sheep, confirmed through comprehensive morphological analysis. To further characterize this isolate and clarify its phylogenetic position among ruminant-infecting Sarcocystis species, we performed molecular characterization by sequencing and analyzing four genetic markers: the 18S rRNA, 28S rRNA, and mitochondrial cox1 genes, and the ITS-1 region.

Individual cysts preserved in sterile distilled water at − 20°C were used for genetic DNA extraction. DNA was extracted from three sarcocysts (each from a different sheep) using the TIANamp Genomic DNA Kit (Tiangen Biotech Ltd., Beijing, China) according to the manufacturer’s instructions. Four genetic markers–18S rRNA, 28S rRNA, ITS-1, and mitochondrial cox1– were amplified from each sarcocyst using the primers specified in Table 1.

PCR was performed in a 25-µL reaction mixture containing: 1X PCR buffer, 0.15 mM MgCl₂, 0.25 mM dNTPs, 1 U of Taq DNA polymerase (TaKaRa, Dalian, China), 50–100 ng of template DNA, and 25 pmol of each primer. Amplification was carried out in a Bio-Rad T100 thermal cycler with the following program: initial denaturation at 95°C for 5 min; 35 cycles of 94°C for 1 min, 53–57°C (primer-specific) for 1 min, and 72°C for 1 min; followed by a final extension at 72°C for 10 min. PCR products were purified using the E.Z.N.A.® Gel Extraction Kit (Omega Bio-Tek, Inc., USA), ligated into the pCE2 TA/Blunt-Zero vector (5 min TA/Blunt-Zero Cloning Kit, Vazyme Biotech Co., Ltd., Nanjing, China), and transformed into Trelief™ 5α Chemically Competent Cells (Tsingke Biotechnology Co., Ltd., Beijing, China). Positive clones were sequenced bidirectionally on an ABI PRISM™ 3730 XL DNA Analyzer (Applied Biosystems, USA).

^aforward primer; ^breverse primer

The obtained sequences were assembled using the SeqMan II program (DNASTAR, USA) based on multiple overlapping regions. Sequence identity and similarity analyses were performed using BioEdit software (Hall 1999). Initial characterization was conducted by comparing the sequences against the GenBank database using the online BLASTn tool (National Center for Biotechnology Information, NIH, USA).

Phylogenetic analyses were conducted separately on the nucleotide sequences of the 18S rRNA, 28S rRNA, and mitochondrial cox1 using MEGA 11 software (Tamura et al. 2021). Reference sequences of Sarcocystis spp. for each gene were retrieved from GenBank. The 18S rRNA and 28S rRNA sequences were aligned using the “R-Coffee” web server, which integrates predicted secondary RNA structure to enhance alignment accuracy (Di Tommaso et al. 2011). The mitochondrial cox1 sequences were aligned using the MUSCLE algorithm embedded in the MEGA 11. All alignments were manually inspected and trimmed at both ends to guarantee uniform start and stop positions across all sequences. The final aligned datasets were as follows: the aligned 18S rRNA nucleotide sequences (24 sequences from 23 species) comprised 1944 positions, ranging from position 60 to 1853 of S. gigantea (MK420020); the aligned 28S rRNA nucleotide sequences (21 sequences from 20 species) consisted of 1831 positions, ranging from position 1 to 1641 of S. gigantea (U85706); and the aligned cox1 nucleotide sequences (22 sequences from 20 species) consisted of 1020 positions, ranging from position 1 to 1020 of S. gigantea (MK420012) without any gaps.

Maximum likelihood (ML) phylogenetic trees were constructed for the 18S rRNA, 28S rRNA, and mitochondrial cox1 sequences using the Tamura 3-parameter, Hasegawa-Kishino-Yano, and Kimura 2-parameter models, respectively. These models were selected based on the lowest BIC (Bayesian Information Criterion) and AICc (Akaike information Criterion, corrected) values determined through ML analysis implemented in MEGA11. All positions containing gaps and missing data were removed using the complete deletion option. The final datasets consisted of 1560 positions for 18S rRNA, 1440 positions for 28S rRNA, and 1020 positions for mitochondrial cox1. The reliability of the ML phylograms was assessed using the bootstrap method with 1000 replications. Toxoplasma gondii and Hammondia spp. were selected as outgroups to root the phylogenetic trees.

Morphological characterization of S. medusiformis sarcocysts

Macroscopically visible sarcocysts (Fig. 1a) were detected in 4 out of 92 sheep (4.3%), exclusively localized in skeletal muscles and diaphragms, with no observed infection in cardiac tissue. Under LM, the sarcocysts measured 2490–4796 µm in length and 248–405 µm in width, exhibiting a thin (1–2 µm thick), striated cyst wall (Fig. 1b). Internally, septa subdivided the sarcocyst into compartments densely packed with banana-shaped bradyzoites measuring 15.3–18.6 × 3.4–4.2 µm (Fig. 1c).

Ultrastructural analysis further revealed that the cyst wall was covered with trapezoidal villar protrusions (VPs), each lined by a distinct electron-dense layer (Fig. 1d, e). These VPs measured 0.8–1.3 µm in length and contained scattered microtubes extending from the apex to the base, without penetrating the underlying ground substance layer (GSL) (Fig. 1e). The GSL was 2.0–2.3 µm thick. Additionally, coiled serpentine filaments were observed originating both the surfaces of the VPs and the intervening areas between them. These ultrastructure characteristics correspond to the type 20 cyst wall as defined by Dubey et al. (2016), confirming the identification of the parasite as S. medusiformis.

Molecular characteristics of S. medusiformis

Sequencing of three S. medusiformis sarcocysts isolates from different sheep yielded complete sequences for 18S rRNA (1928 bp), 28S rRNA (3468 bp), ITS-1 (602 bp) and partial cox1 (1085 bp). All 18S and 28S rRNA sequences showed 100% identity among isolates, while ITS-1 and cox1 exhibited near-identity, with sequences similarities of 99.7–100% and 99.0-100%, respectively. Representative nucleotide sequences have been deposited in GenBank database under the following accession numbers: PV460240 (18S rRNA), PV460246 (28S rRNA), PV470857 and PV470858 (ITS-1), PV468778 and PV468779 (cox1).

Comparison with existing sequences in GenBank (Table 2) revealed that the newly obtained 18S rRNA, 28S rRNA, and cox1 sequences showed the highest similarity (up to 100% identity) to S. medusiformis. The next closest matches were S. gigantea from sheep and S. moulei from goats. In contrast, BLAST analysis of the newly generated ITS-1 sequences showed no significant similarity to any entries currently available in GenBank.

Phylogenetic analysis based on 18S rRNA (Fig. 2a), 28S rRNA (Fig. 2a) and mitochondrial cox1 (Fig. 2c) revealed revealed that the newly sequenced S. medusiformis isolates formed a well-supported clade with S. gigantea and S. moulei. These species infect sheep or goats, produce macroscopically visible cysts, and share felids as their definitive hosts. In the trees inferred from 18S rRNA and cox1 sequences, this S. medusiformis clade grouped with other macrocyst-forming, felid-definitive Sarcocystis species from large ruminants, including S. hirsute from cattle (Bos taurus) and S. buffalonis and S. fusiformis from water buffalo (Bubalus bubalis). In contrast, the topology of the 28S rRNA-based tree differed. Here, the S. medusiformis clade was placed within a larger group comprising Sarcocystis species that form microcysts and utilize canids as definitive hosts. This group included S. tenella and S. arieticanis in sheep, S. capracanis and S. hircicanis in goats, S. cruzi in cattle, and S. poephagicanis in yaks. Notably, the felid-associated, macrocyst-forming species from cattle and water buffalo (S. hirsuta, S. buffalonis, S. fusiformis) formed a basal group to this larger cluster in this particular phylogenetic reconstruction.

Railliet (1886) first described large, ovoid sarcocysts in sheep esophagi, naming them Sarcocystis (Balbiania) gigantea. Later, Mehlhorn and Scholtyseck (1973) provided the ultrastructural description of these sarcocysts, characterized by a thick, double-layered wall with numerous "cauliflower-like" protrusions. Nearly a century after the initial description, Collins et al. (1979) identified a distinct macroscopic sarcocyst in sheep skeletal muscle, distinguished by a thin primary cyst wall (< 2 µm) bearing "snake-like" projections on both the villar and inter-villar surface. Based on these unique morphological traits, it was designated S. medusiformis. In the present study, all observed macroscopic sarcocysts exhibited thin, striated primary cyst walls consistent with the ultrastructure of S. medusiformis originally described in New Zealand sheep (Collins et al. 1979). According to the classification system of Dubey et al. (2016), the cyst wall conforms to type 20.

Sarcocystis species in sheep exhibit a global distribution, with microcyst-forming species (utilizing canids as definitive hosts) being more prevalent than macrocyst-forming species (transmitted via felids) (Dubey et al. 2016; Feng et al. 2023). In China, both microcysts and macrocysts have been detected. Reported prevalence rates for microcysts range from 33.85% to 91.9% (Hu et al. 2017; Dong et al. 2018; Kang et al. 2024), while macrocysts have been reported at 29.1% (Sun et al. 2021). Previous morphological and molecular analyses in Chinese sheep have identified three species: S. tenella, S. arieticanis, and S. gigantea. Our team has previously characterized the microcyst-forming species in detail (Hu et al. 2017); the present study focused on the first identification of S. medusiformis in China. Among the 92 sheep examined, 4 (4.3%) harbored macroscopic cysts of S. medusiformis, marking the first record of this parasite in the country. This prevalence is lower than rates reported in Egypt (5.7%) (El-Morsey et al. 2021), Iran (7.52%) (Farhang-Pajuh et al. 2014), and Italy (12.3%) (Pipia et al. 2016), but higher than that in Australia (3.1%) (Obendorf and Munday 1987). Experimental transmission studies confirm that the domestic cat (Felis catus) is the definitive host for S. medusiformis (Collins 1979; Obendorf and Munday 1987). However, the parasite exhibits relatively low infectivity in cats. Notably, no oocyst or sporocyst shedding was detected in cats fed cystozoites from experimentally infected lambs at 260, 300, and 487 days post-infection (Obendorf and Munday 1987). This reduced infectivity may contribute to the overall low prevalence of S. medusiformis in sheep observed in the current study and in previous reports.

Molecular analysis provides a more sensitive and reliable approach for identifying Sarcocystis species than traditional morphology, particularly given the developmental changes sarcocysts undergo and the existence of morphologically similar cysts in closely related intermediate hosts (Gjerde 2013; Dubey et al. 2016). In this study, we successfully sequenced and submitted 18S rRNA, 28S rRNA, cox1, and ITS-1 sequences of S. medusiformis to GenBank. Notably, the ITS-1 sequences represent the first entry of this genetic marker for the species. BLASTn analysis confirmed up to 100% identity between our sequences and existing S. medusiformis references at the 18S rRNA, 28S rRNA, and cox1 loci, providing robust molecular support for our morphological identification. Phylogenetic analysis consistently placed S. medusiformis within a well-supported clade that includes S. gigantea and S. moulei; however, S. gigantea and S. moulei were more closely related to each other than to S. medusiformis. This phylogenetic pattern reflects the morphological parallels often observed among Sarcocystis species infecting related ruminant hosts. For instance, in sheep, S. tenella, S. arieticanis, and S. gigantea have their morphological counterparts in goats: S. capracanis, S. hircicanis, and S. moulei, respectively (Dubey et al. 2016). Interestingly, while this host-associated pattern is common, sarcocysts morphologically similar to S. medusiformis have not been documented in goats. The sole exception is an unusual finding of similar sarcocysts in an addax (Addax nasomaculatus) in a European zoo (Stolte et al. 1996).

To date, reports on S. medusiformis infections in sheep remain extremely limited, both in case number and geographic distribution. This study provides the first documented evidence of natural S. medusiformis infection in Chinese sheep, significantly expanding the known geographic range of this parasite. Remarkably, sarcocysts morphologically similar to S. medusiformis have been identified only once in a wild bovid species, the addax. The scarcity of available data highlights the need for more extensive sampling of wild and domestic bovid ruminants. Focusing on S. medusiformis and morphologically similar species will enable a deeper understanding of their prevalence, distribution, host specificity, and interspecies relationships within this host group.